Prevalence and Clonal Distribution of Azole-Resistant Candida parapsilosis Isolates Causing Bloodstream Infections in a Large Italian Hospital

Abstract

The most prevalent cause of nosocomial bloodstream infection (BSI) among non-C. albicans Candida species, Candida parapsilosis, may not only be resistant to azole antifungal agents but also disseminate to vulnerable patients. In this survey of BSIs occurring at a large Italian hospital between May 2014 and May 2019, C. parapsilosis accounted for 28.5% (241/844) of all Candida isolates causing BSI episodes. The majority of episodes (151/844) occurred in medical wards. Across the 5 yearly periods, the rates of azole non-susceptibility were 11.8% (4/34), 17.8% (8/45), 28.6% (12/42), 32.8% (19/58), and 17.7% (11/62), respectively, using the Sensititre YeastOne® method. Among azole non-susceptible isolates (54/241; 22.4%), 49 were available for further investigation. Using the CLSI reference method, all 49 isolates were resistant to fluconazole and, except one (susceptible dose-dependent), to voriconazole. Forty (81.6%) isolates harbored the Erg11p Y132F substitution and nine (18.4%) isolates the Y132F in combination with the Erg11p R398I substitution. According to their genotypes, as defined using a microsatellite analysis based on six short tandem repeat markers, 87.7% of isolates (43/49) grouped in two major clusters (II and III), whereas 4.1% of isolates (2/49) belonged to a separate cluster (I). Interestingly, all the isolates from cluster II harbored the Y132F substitution, and those from cluster III harbored both Y132F and R398I substitutions. Of 56 non-Italian isolates included as controls, two Indian isolates with the Y132F substitution had a genotype clearly differing from that of the isolates from clusters II and I. In conclusion, these findings show the dominance of clonal Y132F isolates in our hospital and suggest detection of the Y132F substitution as helpful tool to prevent transmission among hospitalized patients at risk of BSI.

Keywords: Candida parapsilosis; antifungal susceptibility; azole resistance; bloodstream infections; microsatellite genotyping; molecular clones.

Figure 1

Temporal distributions of Candida parapsilosis bloodstream infections (BSIs) in different hospital wards by all (total bar height) and fluconazole or voriconazole non-susceptible (white lines) causative isolates. Overall, BSIs caused by all isolates were 241, whereas those caused by fluconazole non-susceptible isolates were 54 and those caused by voriconazole non-susceptible isolates were 49 (if considering one antifungal drug regardless of the other). As voriconazole non-susceptible isolates were also fluconazole non-susceptible (n = 49), the numbers of voriconazole non-susceptible isolates were coincident (one white-line indicated) or lower (the bottom white-line indicated) than the numbers of fluconazole non-susceptible isolates.

Figure 2

Dendrogram of similarity based on the microsatellite genotypes from 115 Candida parapsilosis isolates. The scale bar in the upper left corner indicates the similarity percentages, whereas the columns on the right report details on the 58 Italian isolates. Clusters I–III in the boxes denote those fluconazole-resistant isolates that harbored an amino acid substitution in the Erg11p, the azole-target enzyme encoded by the ERG11 gene.

Figure 3

Minimum spanning tree analysis based on the microsatellite genotypes from 115 Candida parapsilosis isolates. Each circle corresponds to a unique genotype, whereas the sizes of the circles correspond to the numbers of isolates in the same genotype. The lines between the circles represent relative distance between isolates. The colors of the circles represent in (A) isolates from different hospitals and in (B) isolates harboring or not harboring the Erg11p Y132F (alone or in combination with the Erg11p R398I) substitution.